The goal of this proposal is to elucidate among naturally occurring human malignancies novel molecular defects that inactivate the TGF-beta pathway of tumor suppression. Recent investigations by our group of collaborators have identified the TGF-beta receptor as a new member of the tumor suppressor gene family and have shown that TGF-beta receptor gene mutation is important in the pathogenesis of certain colon cancers. TGF-beta inhibits the growth of many epithelial cell types and for example both abolishes growth of and induces apoptosis in nontransformed colon epithelial cells. In contrast, colon and many other cancer cell lines are resistant to TGF-beta growth suppression. Our recent studies have demonstrated TGF-beta resistance in one group of colon cancers is due to the inactivation by somatic mutation of the type II component (RII) of the TGF-beta receptor complex. Moreover, we have shown that correcting, by RII gene transfer, the RII mutations in such colon cancer cells abolishes the cells capacity to form tumors. The TGF-beta receptor is thus a novel tumor suppressor whose inactivation directly contributes to carcinogenesis. The TGF-beta receptor is composed of a heterodimeric complex of type I (RI) and type II (RII) subunits. In model systems, mutation of either RI or RII inactivates the receptor complex and confers TGF-beta resistance. The hypothesis of this proposal is that most cancers escape TGF-beta growth control by mutation of the RII component of the TGF-beta receptor, mutation of the type RI component of the TGF- beta receptor, or mutation of downstream components of the TGF-beta signalling pathway. We initially identified RII mutations predominantly within a ten base pair polyadenine repeat (BAT-RII) present in the RII coding sequence and also identified these mutations as characteristic only of those colon cancers that also demonstrated the genetic defect of microsatellite instability (RER colon cancers). The RER phenotype is characteristic of colon, endometrial and gastric cancers that arise with high incidence in kindreds with the Hereditary Nonpolyposis Colon Cancer familial cancer syndrome. RER is also present in many cancers of the proximal (right sided) colon, the endometrium and the stomach arising among individuals in the general population who do not belong to cancer families. Recently, we have now identified in some RER colon cancers added novel (not BAT-RII) sites of RII mutation. Most recently, we have also identified among NonRER colon cancer the first case that bears a novel site of RII mutation, and an added case that is the first to likely bear a mutation in post-RII elements of the TGF-beta signalling pathway. The specific aims of this proposal will now determine: i) how these novel mutations functionally inactive TGF-beta signalling; ii) how commonly TGF-beta resistance contributes to carcinogenesis among NonRER colon cancers; iii) how often TGF-beta resistance in NonRER colon cancers is mediated by RII mutations, by RI mutations, or by defects in post receptor components of the TGF-beta signalling pathway; iv) the role of BAT-RII mutations in the genesis of RER cancers characteristic of certain noncolonic tissues (stomach and endometrium), as well as whether NonRER gastric and endometrial cancers, like NonRER colon cancers, bear additional novel mutations that inactivate TGF-beta signalling. Lastly, we will determine whether a diagnostic test for BAT-RII mutations will allow early detection in man of RER colon cancers.